Answers are grounded in this paper and cite their sources.
Generative artificial intelligence (GenAI) is increasingly entering universities not only as a student-owned productivity tool but as part of the digital infrastructure through which learning support, feedback and assessment guidance are mediated. This concise paper reports on an exploratory pilot of Cogniti AI agents in an Aotearoa New Zealand foundation business course, with a focus on human-centred learning design. The pilot involved five role-specialised agents: a course buddy for administration support, a case-discussion agent for Maori entrepreneurship learning, and three virtual team members focused on task coordination, relationship management and value-adding contribution. Rather than positioning AI as a generic chatbot, the paper presents a design model for embedding AI agents within the course ecosystem. Drawing on design records, early usage patterns and practice reflections, the paper argues that AI agents can extend human-centred pedagogy when they are co-created with students, shaped by student input, assessment-aligned, and circumscribed by teacher oversight. The contribution is a practical model for integrating AI agents into learning management systems in ways that prioritise connection, mentorship, evaluative judgement, feedback literacy and collaborative learning.
No takes yet. Share an insight, caveat, or question.
Pillay et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: